Python 如何在returnn中的新网络中加载经过训练的网络的某一层的权重？

我在文件夹路径/to/modelFile中有以下网络的训练权重：

network={
"conv_1" : {"class": "conv", "filter_size": (400,), "activation":"abs" , "padding": "valid", "strides": 10, "n_out": 64 },
"pad_conv_1_time_dim" : {"class": "pad", "axes": "time", "padding": 20, "from": ["conv_1"]},
"conv_2" : {"class": "conv", "input_add_feature_dim": True, "filter_size": (40, 64), "activation":"abs", "padding": "valid","strides": 16, "n_out": 128, "from": ["pad_conv_1_time_dim"]},
"flatten_conv": {"class": "merge_dims", "axes": "except_time","n_out": 128,  "from": ["conv_2"]},
"window_1": {"class": "window", "window_size": 17, "from": ["flatten_conv"]},
"flatten_window": {"class": "merge_dims", "axes":"except_time","from": ["window_1"]},
"lin_1" :   { "class" : "linear", "activation": None, "n_out": 512,"from" : ["flatten_window"] },
"ff_2" :   { "class" : "linear", "activation": "relu", "n_out": 2000, "from" : ["lin_1"] },
"output" :   { "class" : "softmax", "loss" : "ce", "from" : ["ff_2"] }
}

我想将层“conv_1”和“conv_2”的训练权重加载到以下网络中：

network={
"conv_1" : {"class": "conv", "filter_size": (400,), "activation": "abs" , "padding": "valid", "strides": 10, "n_out": 64 },
"pad_conv_1_time_dim" : {"class": "pad", "axes": "time", "padding": 20, "from": ["conv_1"]},
"conv_2" : {"class": "conv", "input_add_feature_dim": True, "filter_size": (40, 64), "activation":"abs", "padding": "valid", "strides": 16, "n_out": 128, "from": ["pad_conv_1_time_dim"]},
"flatten_conv": {"class": "merge_dims", "axes": "except_time", "n_out": 128,  "from": ["conv_2"]},
"lstm1_fw" : { "class": "rec", "unit": "lstmp", "n_out" : rnnLayerNodes, "direction": 1, "from" : ['flatten_conv'] },
"lstm1_bw" : { "class": "rec", "unit": "lstmp", "n_out" : rnnLayerNodes, "direction": -1, "from" : ['flatten_conv'] },
"lin_1" :   { "class" : "linear", "activation": None, "n_out": 512, "from" : ["lstm1_fw", "lstm1_bw"] },
"ff_2" :   { "class" : "linear", "activation": "relu", "n_out": 2000, "from" : ["lin_1"] },
"ff_3" :   { "class" : "linear", "activation": "relu", "n_out": 2000,"from" : ["ff_2"] },
"ff_4" :   { "class" : "linear", "activation": "relu", "n_out": 2000,"from" : ["ff_3"] },
"output" :   { "class" : "softmax", "loss" : "ce", "from" : ["ff_4"] }
}

---

这在Return怎么可能

使用

子网络层

是一种选择。这看起来像：

trained_network_model_file = 'path/to/model_file'

trained_network = {
"conv_1" : {"class": "conv", "filter_size": (400,), "activation": "abs" , "padding": "valid", "strides": 10, "n_out": 64 },
"pad_conv_1_time_dim" : {"class": "pad", "axes": "time", "padding": 20, "from": ["conv_1"]},
"conv_2" : {"class": "conv", "input_add_feature_dim": True, "filter_size": (40, 64), "activation":"abs", "padding": "valid", "strides": 16, "n_out": 128, "from": ["pad_conv_1_time_dim"]},
"flatten_conv": {"class": "merge_dims", "axes": "except_time","n_out": 128,  "from": ["conv_2"]}
}

network = {
"conv_layers" : { "class" : "subnetwork", "subnetwork": trained_network, "load_on_init": trained_network_model_file, "n_out": 128},
"lstm1_fw" : { "class": "rec", "unit": "lstmp", "n_out" : rnnLayerNodes, "direction": 1, "from" : ['conv_layers'] },
"lstm1_bw" : { "class": "rec", "unit": "lstmp", "n_out" : rnnLayerNodes, "direction": -1, "from" : ['conv_layers'] },
"lin_1" :   { "class" : "linear", "activation": None, "n_out": 512, "from" : ["lstm1_fw", "lstm1_bw"] },
"ff_2" :   { "class" : "linear", "activation": "relu", "n_out": 2000, "from" : ["lin_1"] },
"ff_3" :   { "class" : "linear", "activation": "relu", "n_out": 2000, "from" : ["ff_2"] },
"ff_4" :   { "class" : "linear", "activation": "relu", "n_out": 2000, "from" : ["ff_3"] },
"output" :   { "class" : "softmax", "loss" : "ce", "from" : ["ff_4"] }
}

我想对你来说，这是我的首选

否则，每个层都有

custom_param_importer

选项，您可以让它使用该选项

---

然后，对于许多层，您可以为参数定义初始值设定项，例如，对于

ConvLayer

，您可以使用

forward\u weights\u init

。可以使用

load\u txt\u file\u初始值设定项之类的函数，或者可以添加类似的函数直接从TF检查点文件加载。
如果子网络中缺少“输出层”错误，可以将名称“flattle\u conv”更改为“output”它应该工作如果子网络的重量不应该被训练，参数{…，“trainable”：False，…}必须被添加到训练网络中包含可训练参数的所有层以及子网络层“conv_layers”